Antisense therapy requires the use of modified nucleotides in order to block gene expression by preventing messenger RNA (mRNA) translation into active proteins. The criteria requested for optimal activity of an antisense oligonucleotide are resistance to endo- and exonucleases, affinity and specificity for the targeted mRNA sequence as well as easy uptake by the concerned cells, and finally destruction or inactivation of the mRNA. Certain chemical modifications of the nucleotides or of the phosphate backbone can allow such biological activity and several approaches have been used to this end.
The first-generation antisense oligonucleotides consist of DNA in which a sulfur atom has been incorporated in replacement of an equatorial oxygen atom in the phosphate backbone. This PS backbone confers a better resistance to nucleases, although it lowers the affinity of the oligonucleotide for its target. The mechanism of action involves RNase H, an enzyme that specifically degrades the RNA in a DNA:RNA complex, therefore destroying the mRNA. The second-generation strategies use mixtures of modified oligoribonucleotides and oligodeoxyribonucleotides in chimeric molecules. The part of the molecule composed of ribonucleotides is then responsible for the affinity for the targeted sequence, whereas the part containing the deoxyribonucleotides acts as a substrate for RNase H which cleaves the targeted mRNA. The ribonucleotides are modified at the C2xe2x80x2 position of the sugar where different substituents can be attached.
Chemical synthesis of the 2xe2x80x2-modified nucleotides is currently very expensive and, as the demand for such compound is expected to soar in the near future, their biosynthesis by an appropriate microorganism would markedly lower the cost of their production, especially if it is possible to start with cheap compounds available in large amounts.
Cyclic nucleotides are essential elements in many biochemical processes of the eukaryotic and prokaryotic cell cycles, particularly adenosine 3xe2x80x2,5xe2x80x2 cyclic monophosphate and guanosine 3xe2x80x2,5xe2x80x2 cyclic monophosphate. These compounds constitute the most adequate starting material for the purpose of the present invention.
It is known that such natural cyclic nucleotides could be transformed to their corresponding 5xe2x80x2-nucleoside monophosphates by 3xe2x80x2,5xe2x80x2-cyclic nucleotide phosphodiesterases e.g., from Vibrio fischeri (WO 94/16063, Journal of Bacteriology, Vol. 175, No. 15, p. 4615-4624). Furthermore the conversion of 2xe2x80x2 O-methyl- and 2xe2x80x2 O-ethyl- nucleoside 3xe2x80x2,5xe2x80x2-cyclic phosphates to the corresponding 5xe2x80x2-phosphates by enzymatic hydrolysis has been published (Biochemistry, Vol. 11, No. 26, 1972, p. 4931-4937).
There is still a need for a process which allows the synthesis of nucleotides with more complex 2xe2x80x2-substituents.
Surprisingly an enzymatic process has now been found to produce nucleosides with such more complex 2xe2x80x2-substituents.
The invention relates to a process comprising treatment of compounds of Formula (I) 
wherein
B is adenine, guanine or hypoxanthine
Z is hydrogen or a negative charge
R is xe2x80x94[CH2CH(R1)xe2x80x94O]nxe2x80x94R2, xe2x80x94CH2CH2X, in which
R1 is hydrogen or (C1-C6) alkyl
R2 is hydrogen or (C1-C6) alkyl
n is a number from 1 to 6
X is OH, F, NR3R4 
R3 and R4 are independently from each other hydrogen or (C1-C6) alkyl with a biocatalyst having cyclic phosphodiesterase activity.
The product of such process may in a second step be hydrolysed to a compound of Formula (II). 
wherein B and R have the same meaning as defined above. This hydrolysis could be done chemically or enzymatically.
Biocatalyst having cyclic phosphodiesterase activity may be for the purpose of this invention a purified enzyme, from original or recombinant source, a partially purified enzyme, a crude cell extract, enzymes immobilized or caught in partially broken cells or cell debris. The biocatalyst may be used in solution, suspension or in immobilized form (matrix bound) according to routine processes, e.g. batch or continuous, known in the art. Source for the biocatalyst may be every organism, e.g. animal organs, microorganisms like fungi or bacteria, which comprise an enzyme suitable for being used as biocatalyst in the above process. Whether a microorganism contains such enzyme can easily be evaluated by a test which comprises the comparative growth of an organism on a medium containing only salts and Tris buffer (minimal medium, described in Dunlap et al. J. Gen. Microbiol. 1992, 138, 115-123) and in parallel on the same medium supplemented with 1 to 5 mM of compound of formula (I). Any microorganism containing an extracellular or periplasmic (for gram-negative bacteria) enzyme able to hydrolyse the substrate will grow on the supplemented medium. The minimal medium serves as a negative control differentiating between microorganisms able to grow on Tris or on cAMP only, as some microbes are known to be able to utilise Tris as a nutrient. For microorganism possessing a cytoplasmic phophodiesterase (as it is the case for many bacteria, e.g., Escherichia coli or Salmonella typhimurium), growth on nutrient agar is necessary. The cell disruption supernatant has to be assayed as the cytoplasmic membrane is known to be fairly impermeable to cyclic nucleotides.
Accordingly the present invention also comprises new enzymes which can be identified in the above tests.
A preferred embodiment of the invention is the use of the gram negative bacterium Serratia marcescens, particularly the use of strain DSM 30121 as the source of biocatalyst having cyclic nucleotide phophodiesterase activity. Such biocatalyst having cyclic nucleotide phosphodiesterase activity is used as a purified enzyme, a partially purified enzyme or as a crude cell extract.
For the specific cleavage of the 3xe2x80x2,5xe2x80x2 nucleoside cyclic monophosphate to the 5xe2x80x2-monophosphate only the enzymatic cleavage is selective and therefore clearly superior to chemical hydrolysis.
Furthermore the invention relates to compounds of formula (I) 
wherein
B is adenine, guanine or hypoxanthine
Z is hydrogen or a negative charge
R is xe2x80x94[CH2CH(R1)xe2x80x94O]nxe2x80x94R2, xe2x80x94CH2CH2X, in which
R1 is hydrogen or (C1-C6) alkyl
R2 is hydrogen or (C1-C6) alkyl
n is a number from 1 to 6
X is OH, F, NR3R4 
R3 and R4 are independently from each other hydrogen or (C1-C6) alkyl
Preferred are compounds of formula (I) wherein R is CH2CH2OR2, preferably CH2CH2OCH3 further preferred are compounds wherein B is adenine.
The compounds of formula (I) could be synthesized by reacting the corresponding nucleoside 3xe2x80x2,5xe2x80x2 cyclic monophosphates with an alkylation reagent, for example R-halogene, especially R-Br or R-I.
Furthermore the invention comprises the use of a biocatalyst having cyclic phosphodiesterase activity for the preparation of compounds of formula (II).